Discourses in Residential Child Care and Possibilities for Evidence-Based Practice

This article explores professional discourses in the Norwegian residential child care system. It discusses how the discourses serve as constraints on and possibilities for evidence-based practice when different definitions of evidencebased practice are considered. Among the Nordic countries, Norway has been a forerunner in the implementation of evidence-based practice in child welfare. However, I argue that tensions exist, both within professional practice and between professional understandings and policy aims. I use discourse theory to analyze interviews with 19 professionals working in coercive residential child care. The results reveal two competing professional discourses: the discourse of technoscience and the discourse of indeterminacy. Possibilities of evidence-based practice in residential child care are found within both discourses if a wide and inclusive definition of evidence-based practice is applied. This study emphasizes the importance of engaging in constant reflection when discussing possibilities for evidence-based practice within residential child care.

Over the past decade, evidence-based practice has become the dominant paradigm in European child welfare services (Grietens, 2013).Evidence-based practice can be interpreted as a coupling between policy and practice, in which ideas of management and steering are combined with scientific knowledge so as to increase the effectiveness of services (Bergmark & Lundström, 2006;Foss Hansen & Rieper, 2009).However, evidence-based practice has been characterized as a contested concept with regard to its content and validity for implementation into professional practice (Backe-Hansen, 2009;Barfoed & Jacobsson, 2012;Bergmark & Lundström, 2006;Mullen & Streiner, 2006;Satterfield et al., 2009).Several researchers have pointed out the lack of consensus in the practice field concerning the epistemological foundations of evidence-based practice.In addition, researchers have frequently questioned what counts as correct knowledge and how to obtain it (Angel, 2003;Axford & Morpeth, 2013;Ekeland, 1999;Gilgun, 2005;Grimen & Terum, 2009;Webb, 2001).
One way of dealing with the contested concept of evidence-based practice has been to use two types of definitions-a narrow definition and a wide one (Backe-Hansen, 2009).A narrow definition builds on a hierarchy, with knowledge that can be derived from randomized control trials (RCTs) being defined as the gold standard for decisions regarding professional practice (Backe-Hansen, 2009).A wide definition typically integrates best available research evidence with professional expertise and patient preferences when clinical decisions are made (Backe-Hansen, 2009).This wide definition has also been referred to as the "three-circle model of evidence-based clinical decisions" (Satterfield et al., 2009, p. 371).
This article aims to explore professional discourses in the Norwegian residential child care system and discusses their possibilities for accommodating the different definitions of evidence-based practice.Coercive residential child care for juveniles with serious drug or behavioral problems is used as the empirical case.
By engaging in conceptual discussions and addressing criticisms directed at the concept of evidence-based practice, researches have shed light on the challenges of implementing evidence-based practice into social work (Backe-Hansen, 2009;Mullen, Shlonsky, Bledsoe, & Bellamy, 2005;Mullen & Streiner, 2006).Mullen and colleagues (2005), drawing on discussions in the literature, define eight challenges that need to be resolved if evidence-based practice is to become a reality in social work (p. 67).Based on an empirical study, Barfoed and Jacobsson (2012) explore the launching of one specific evidence-based assessment instrument (the ASI interview) into the Swedish social services, and question whether it has been fully accepted and institutionalized.They highlight the need of more empirical research on social work practice.For instance, what is included or excluded when social workers collect "facts" about clients, or what assumptions of a moral or decent life are embedded in social work practice (p.16).The present study aims to contribute to this research field by empirical research about social workers' own understanding of practice and how it relates to different ideas of evidence-based practice.
A recent Norwegian study concludes that work in residential child care institutions is primarily characterized by experience, intuition, and feelings.This finding poses a managerial challenge with regard to the intention of policymakers to implement new (evidence-based) programs and manuals.The authors ask if it is possible to integrate both the use of explicit knowledge and the use of experimental knowledge in social work practice (Gotvassli, Augdal & Rotmo, 2014).The present article further explores this question.
I suggest that how evidence-based practice is defined will influence its possibilities of being accommodated within the professionals' understandings of social work practice.The commitment of the actors directly involved in the implementation process is perceived as an important factor for successful implementation (Howlett & Ramesh, 2003).Hence, the professionals' own understandings of social work practice will be important when discussing the conditions for evidence-based social work.By exploring professional discourses, the present study adds to the body of knowledge on the possibilities and challenges for evidence-based social work, in particular social work in the residential child care system.
This article builds on a qualitative study of 19 professionals working in different residential care units.I ask two main research questions: What are the discourses in use when professionals give meaning to their practice?What are the constraints on and possibilities for evidence-based practice found within the identified discourses when different definitions of evidence-based practice are considered?
The inspiration behind this article is poststructural discourse theory, and in the analysis of professional discourses, I use parts of a theoretical framework developed by Laclau and Mouffe (1985).This framework provides insight into the processes in which meaning is formed and the practical implications it gains.By interpreting the way in which social workers use talk to make meaning of their practice, one can gain insight into dominant discourses.Discourses organize how one understands reality, and an important precondition of discourse theory is the existence of several competing discourses at one time, making possible contradictory understandings of the world.Discourses organize what can be included and what can be excluded from one's horizon of meaning.An exploration into the professional discourses in residential child care is therefore a fruitful approach for discussing how social workers' understandings of practice can represent both possibilities for and constraints on evidence-based practice within residential child care.
In my discussion of professional discourses and their relation to evidence-based practice, I also use perspectives from the sociology of professions-focusing on the tensions between formal knowledge, discretion, and autonomy, which are important elements of a profession's legitimacy and position (Freidson, 2001;Jamous & Peloille, 1970;Larson, 1977).The social work profession has been described as having a low status in the professional hierarchy owing to, among other things, its vague theoretical framework and its multiplicity of practices (McDonald, 2003;Molander & Terum, 2008).I discuss how these characteristics of the social work profession can be a source for understanding the discursive formation and, hence, must be considered when discussing the conditions for evidence-based practice.
I begin by describing the state of evidence-based practice in the Norwegian residential child care system.Next, I explain the methodology and the poststructural framework used in this article, the main focus of which is to explore social workers' understandings of practice.The results reveal two competing professional discourses.I then discuss how the discourses relate to different definitions of evidence-based practice.In the concluding section, I suggest a wide and inclusive model of evidencebased practice in residential child care so as to encompass elements from both discourses.
A comparative study from 2009 concludes that there is no doubt that evidencebased practice has gained entry into the Norwegian residential child care system.Among the Nordic countries, Norway seems to be taking the lead in developing and implementing new evidence-based methods for the treatment of drug-and behaviorrelated problems among juveniles (Bengtsson & Böcker Jacobsen, 2009, p. 250).Although evidence-based practice has garnered increased attention in the political agenda in Norway, the social workers' responses to this practice have been varied (Gotvassli et al., 2014).
Norway's implementation of policy initiatives has stimulated discussions not only about the content of professional practice (Backe-Hansen, 2009;Bengtsson & Böcker Jacobsen, 2009;Bergmark & Lundström, 2006;Storø et al., 2010) but also about professional autonomy (Bergmark & Lundström, 2006).On the basis of a theoretical discussion, Backe-Hansen (2009) has suggested using a wider or more inclusive definition of evidence-based practice so as to encompass the nature of child welfare work.She has specifically referred to a definition put forth by Gilgun (2005, p. 52), which takes the following into consideration: research and theory; what professionals have learned from their clients; professionals' personal assumptions, values, biases, and worldviews; and what clients bring to practice situations (i.e., the four cornerstones of evidence-based practice).The rhetorical battle concerning the content of evidence-based practice raises important questions, for example, about the relationship between formal knowledge and discretion in professional work.The content of evidence-based practice has been defined in many different ways, and in the discussion section of the article I will discuss how the use of a wide and a narrow definition of this concept has implications for the possibilities of implementing evidence-based practice in residential child care.

Methodology
This study uses qualitative methods, including a combination of individual in-depth interviews and focus groups.In choosing the methodological framework, I was guided by the poststructural approach in which the study of articulatory practices is defined as important for the analysis of discursive formation and change (Laclau & Mouffe, 1985).Laclau and Mouffe (1985) define articulatory practice as "any practice establishing a relation among elements such that their identity is modified as a result of the articulatory process" (p.105).For the purposes of the present study, articulatory practice is defined as the way in which social workers use talk to make meaning of their practice, particularly the way in which they make use of comparisons and oppositions in this talk.Conducting individual interviews and focus groups enabled me to explore the social workers' subjective understandings of their professional practice and the meaning they attributed to this work (Kvale, 1996;Rubin & Rubin, 2005).
The informants were recruited from five residential care units housing juveniles placed in coercive treatment, as governed by the Child Welfare Act of 1992 (1992).Children with comprehensive drug or behavior problems can be placed in a training or treatment institution by the county social welfare board for up to 12 months without their consent or the consent of those who have parental responsibility for them (Child Welfare Act, 1992).Of the 982 children who were placed in residential care during the last four months of 2014 (Barne-Ungdoms og Familiedirektoratet, 2015), 184 children were placed in coercive residential care (according to statistics from the Norwegian Directorate for Children, Youth and Family Affairs).All the institutions in this study were open institutions (unlocked), which is the usual practice in Norway (Bengtsson & Böcker Jacobsen, 2009).Residential child care in Norway is publicly funded, but a majority of the residential care units are privately owned (Barne-Ungdoms og Familiedirektoratet, 2015).In this study, informants from both private and public care units are represented.
I intentionally selected the residential child care units and the informants for this study so as to ensure variation in the informants' experiences with coercive residential child care and in the informants' educational background.The informants' professional background varied from no education to a master's degree, with most informants holding a relevant bachelor's degree (social worker or child welfare officer).Some of the informants had experience with the new programs and manuals described in the introductory section of the present article, whereas other informants had limited experience with them.The informants' educational background and experiences, however, are not addressed further here because these topics are beyond the scope of this article.In three of the care units, the manager or responsible milieu therapist recruited the informants.In the other two units, the recruitment was carried out by the Regional Office for Children, Youth and Family Affairs.The informants were informed that participation in the study was voluntary, and that they might withdraw their consent as long as the project was in progress.The study was conducted in accordance with ethical guidelines and regulations and was approved by the Data Protection Official for Research (Norwegian Social Science Data Services [NSD]).
I conducted 12 individual interviews, each lasting approximately an hour, at the residential care units.The semi-structured interviews followed a thematic guide.The informants were asked to talk about their daily work, work role, understanding of treatment, knowledge, and purpose of work.Because I was interested in identifying professional discourses, and not the professionals' opinions about policy ideas, I did not ask the informants specifically about evidence-based practice.In addition to the individual interviews, I conducted two focus groups (one with four informants and the other with six informants) to facilitate discussions among the professionals, with the aim of further elaborating on themes that I had become aware of during the individual interviews (Bloor, 2001).Three of the informants in the focus groups were also among those 12 informants who had been interviewed individually.Each focus group session lasted two hours and included a short break.Both the interviews and the focus groups were recorded and transcribed verbatim.
When I analyzed the data gathered from the interviews and the focus groups, I was inspired not only by poststructural discourse theory but also by Haavind's (2000) understanding of interpretative methodology, wherein she describes analysis as occurring in two (or more) rounds.My first step in the analysis was to organize the material according to themes in the interview guide.Further interpretation of meaning into these categories led to the identification of new analytical categories illustrating patterns in the informants' descriptions of their practice.This stage corresponds to what Haavind (2000) describes as the "second round".At this stage, I was specifically interested in interpreting the articulatory practices.I scrutinized the words the informants used to give meaning to their practice and searched for contradictions and oppositions.
The new analytical categories consisted of oppositions-for example, formal knowledge versus discretion, measurable objectives versus subjective valuations.The oppositions reflected the tensions in the informants' descriptions of practice.In the next stage, I searched for discourses that could unify the content within the oppositional categories.I explored how the professionals' understandings of practice could be related to more general ideas of professional work and to recent policy initiatives, thereby identifying professional discourses in use.A precondition in poststructural discourse theory is the existence of wider fields of social meaning and discourses competing for dominance (Howarth, 2005, p. 321).

Findings: Tensions in informants' understandings of practice
In the following subsections, excerpts from interviews conducted with eight of the informants are presented so as to illustrate the informants' understandings of professional practice.Interview excerpts are chosen from these eight informants because they were the ones who most clearly described the tensions related to understandings of professional practice.All informants are anonymized (including gender) and given fictitious names.

Talking about knowledge and methods
The interviews showed variances between informants who preferred to talk about treatment programs and manuals and those who preferred to talk about personal experiences, personalities, and different method combinations.Knut, one of the informants, described, in a detailed and engaged manner, the manual-based method with which he was familiar.Knut's description of the methodological framework was peppered with technical terms, such as "multisystem," "multiple teams," and "institutionalized."When talking about the method, Knut referred to "theories that display which methods are useful and which methods are not."The comparison he made between "useful" and "not useful" indicates that Knut was convinced of this method's effectiveness.Knut offered the following explanation of the advantages of the new approach: This institution aims to use a multisystem approach that makes use of multiple teams.This contrasts with other institutions in which one team does it all.…This is not the juveniles' home.This is a place for short-term treatment.Research shows that juveniles who remain institutionalized for over a year exhibit certain behaviors and struggle to return to normal family relationships afterward.Thus, this institution focuses on short-term treatment.
In his explanation, Knut used both the methodological framework and references to research when comparing a standardized approach with other types of practices.He described the care unit as a place for treatment, not for living.By making this distinction between treatment and living, he opposed an approach to residential child care in Norway that involves long-term treatment and the establishment of intimate relationships (Hassel Kristoffersen et al., 2011, pp. 36-38).In the interview, Knut pointed to scientific knowledge as a better approach to professional decisions than interventions based on subjective emotional responses, as seen in families.Furthermore, he highlighted the importance of learning about the knowledge base and gaining confidence in how to practice the methods.
Sigrid was another informant who emphasized the usefulness of a specific manual-based treatment method.This emphasis is evident in the following interview excerpt: I think it is great.And many juveniles who come here have lots of aggression and cannot channel their anger in a proper way… and ART is really good for that.And many can struggle with social skills.So it is a very useful tool.I think the juveniles do not think it is so useful though.They think that there is lots that they already know.
Sigrid seemed to value having a tool that is directed toward specific behavioral problems.However, unlike Knut, Sigrid did not use technical jargon when she talked about the method.She instead related her understanding of practice to her own personal thoughts and to her understanding of the juveniles' thoughts.Sigrid described ART as being a good technique for aggression management, but she was aware that the juveniles did not find the method to be useful.This combination of enthusiasm and ambivalence was notable in the interview with Sigrid, particularly when Sigrid made the following statement: "Sometimes I think that there is a fine balance between reinforcement and bribery.…It is kind of borderline.…But at the same time, the fact is that it does have an effect-that it works."Sigrid talked about a detailed system of reinforcement that is an integrated part of one of the methods and compared the system of rewards with bribery.At the same time, she justified using the method by noting that it has "an effect." Jon, another informant, highlighted the importance of having the appropriate personal qualifications and knowledge of human interaction for professional practice.When talking about professional practice, Jon refrained from referring to theories or methods.He instead offered the following explanation as to why not everyone can learn to work in residential child care: You work with people in a way that is unlike a job.It is a lifestyle.According to Jon, personality and involvement are important when working with people.He then contrasted personality and involvement with "what you have learned," which can be interpreted as formal knowledge.Anita, another informant, highlighted how possessing knowledge of different methods is an advantage.When a variety of options are presented, the juveniles themselves can experiment to see which method works best."One method will probably not fit all," she said.Her comment does not lend support to the standardization of practice.
According to Anita and Jon, knowledge cannot be derived from theories and methods and applied to individual juveniles without being altered so as to fulfill individual needs and preferences.The professionals positioned themselves as important translators in this process, and the way they talked about the combination of education, personal life experiences, and personal qualifications pointed to discretion as an important component of professional practice.Discretion was not explicitly discussed by the informants but rather was implicitly comprehended as an aspect of practice in their understandings.Discretion is generally understood as the aspect of professional work that allows professionals to use their own judgment in applying general knowledge to particular cases.Hence, professional autonomy is required (Freidson, 2001;Lipsky, 1980;Schön, 1983).
Some of the informants emphasized the importance of exercising professional autonomy in decisions about applying knowledge and methods.Taking a different stance, Olav, the manager of one of the units, stated the following: I feel that we have been continually yearning.Now we have a tool!And the employees say, "Yes!We are on board!"So feedback was completely positive.The time was so right!We have to have something specific.And this method is very specific.It is as if everyone has to think the same way.
Olav talked about one of the (evidence-based) methods referred to in the introductory section of this article.When Olav exclaimed, "Now we have a tool!" he indicated that, unlike the current situation, tools were not available in the past.Like Sigrid, Olav used the word "tool," which can be interpreted as representing an understanding of professional practice that is in need of technical means.In addition, Olav stated that "we have to have something specific."He used the word "specific" when speaking about the new method.He also related the new method to standardization and employee loyalty when he commented that "it is as if everyone has to think the same way."Olav's satisfaction with a unitary approach can be related to his position as a manager (he was the only manager in the study).However, it is interesting that he also described all his employees as being enthusiastic about using this approach.
Therese, another informant, advocated taking a much different approach than the one emphasized by Olav.She perceived not being overly focused on using one particular method as a strength in treatment.In the focus group discussion, she talked about the methods' relative impact on treatment compared with other aspects of treatment: I think that the fact that we are not overly focused on one method is a strength.Because if you look at what actually works and is important in treatment, methods do not mean much-something like 10% or 15%.And then there are the juveniles' own resources, and then there are relationships.In a way, they are the two biggest [factors], the two most important.
According to Therese, both the juveniles' own resources and the establishment of good relationships are more important than the application of specific treatment methods in professional practice.By stating that "we are not overly focused on one method," Therese also shows that she is aware of policy initiatives directed toward (methodological) standardization but that she does not believe it to be consistent with her understanding of practice.She refers to research results ("what actually works") to legitimize a counterapproach.
As revealed in the interview excerpts, the informants sometimes expressed opposing views with regard to several aspects of knowledge and methods in professional practice.Tensions were identified between old versus new practices, what is useful versus what is not useful in treatment, scientific knowledge versus discretion, professional autonomy versus conformity, and methodological standardization versus methodological pluralism.

Talking about work objectives and outcomes
When the informants talked about work objectives and outcomes, tensions developed along the same dimensions.Olav was one of the informants who emphasized the importance of having clear objectives.He was happy to have finally gained a specific work objective: "The goal is, of course, abstinence.This is obvious and clear here.It is very good that we have gained such a clear goal."In stating that "we have gained such a clear goal," Olav appeared to favor the new practice over the old and to suggest that goals had been vague before the introduction of the new method.In the interview, Olav did not explain why he found establishing clear objectives to be important for practice.However, clear objectives can be interpreted as leading to both a reduction in uncertainty in professional decisions and an increase in the possibility of measuring treatment outcome.
In the following interview excerpt, Knut describes one of the standardized methods, highlighting the statistical outcomes: The theory is very detailed, very complicated.I do not know the whole theory yet.I have been here for only five months.It is extensive, and there are often things I question or ask others about.But it seems like it works for many.During the course, they stated that there was a 30% to 50% success rate at regular institutions and a 60% to 80% success rate at these [evidence-based] institutions.So, if you look at the statistics, it appears that this is a method to consider in the future.
Knut's use of the term "success rate" can be related to an understanding of professional practice in which treatment outcomes can be measured in quantifiable terms based on the achievement of specific objectives.In the interview, Knut referred to both statistics and his own experiences when speaking about treatment outcomes.He spoke about being skeptical of the method in the beginning and then becoming more convinced of the method's potential after personally experiencing changes in the juveniles' behaviors.Neither Knut nor Olav related their understandings of outcomes to evaluations or feedback from the juveniles.Instead, they discussed outcomes as things that can be deduced from quantitative indicators of treatment progress combined with personal experiences.
Understandings of practice that differed from the emphasis on objectivity and measurable outcomes were also identified.An informant named Alf described the objective of residential child care is to provide juveniles with a general training for life."This is training for how people treat each other in daily life within society's laws and rules and how we achieve what we understand as a good life," he said.Alf's use of the expression "a good life" indicates an understanding of treatment objectives as being comprehensive and complex.Alf spoke about how his role involved helping juveniles not only to cope with addiction or behavioral problems but also to follow law and order, thus equipping them with the skills needed to adjust to society's rules and expectations.The expression "a good life" is a subjective and normative utterance that makes it difficult to measure the outcome of a treatment in quantifiable terms.
Several informants highlighted that to capture the complexity encountered in their professional work, knowledge of treatment outcome should be related to the juveniles' individual challenges and subjective experiences.In the focus group discussion, Nina identified drug abuse and behavioral deviance as aspects of a more complex totality: "They are sent here for substance abuse treatment, but we see that in many, many cases, substance abuse is the least of their problems.It is often merely a symptom of other challenges".Nina used the word "symptom" to describe the relative impact of the drug problem in relation to other challenges in the juveniles' lives.Nina's description of the juveniles' problems calls for other and more complex treatment objectives than merely "abstinence", as emphasized by Olav.
Therese pointed to the challenges of finding reliable ways to document and legitimate professional work that diverges from the dominant policy: So, I think that [finding] a balance between being a professional and [being] in a relationship while simultaneously fighting the outside world can be a challenge.But it is clear that the biggest challenge for us is to endure the recovery process.And this takes time.…[and a belief] that what happens now will give results later.In recent years, I have felt that the biggest challenge has been to communicate this to the outside world.
Therese was concerned about the increased focus on short-term treatment in some of the new treatment programs, and she talked about treatment as a process.Her use of the word "process" indicates that she considers treatment to be a time-consuming activity rather than an activity that can be carried out in the short term.Following Therese's line of reasoning, this also has consequences for how and when treatment results can be evaluated.Unlike the informants who spoke about outcomes being measured during and at the end of treatment, Therese spoke about having the patience to attain "results later", which makes outcomes difficult to measure.When Therese mentioned "fighting the outside world," she was drawing a distinction between her understanding of practice and the outside world.Therese described obtaining approval for her understanding of "later results" in professional practice to be a challenge.
The interview excerpts presented here show how the informants used certain words and expressions when talking about their understandings of knowledge and methods.Tensions were identified between complex goals versus specific goals, measurable objectives versus subjective valuations, treatment as a process versus treatment as a short-term intervention, and the outside world versus professional understandings.The patterns associated with the informants' understandings of professional practice are interpreted here as being made possible by available and competing discourses.

Discussion
This study revealed the existence of tensions between informants who exhibited an enthusiasm for the new methodological framework and those who emphasized other elements and understandings of practice relating to a more traditional view of professional work.My analysis of the informants' descriptions of their understandings of practice showed that the informants' held opposing views in certain areas-for example, with regard to the past and to competing understandings of residential child care.By conducting a closer examination of the opposing views constructed and their relations to dominant policy ideas and to more general ideas of professional work, I identified two opposing discourses: the discourse of technoscience and the discourse of indeterminacy.
The discourse of technoscience unifies talk about specific goals, measurable objectives, useful methods, standardization, and conformity.The belief in the potential of scientific knowledge in developing universal and effective models for treatment is interpreted as being important for the discursive formation.The policy of quality improvement through increased regulations and guidelines of professional practice is interpreted as being comparable with ideas that are constitutive for the discourse of technoscience.
By contrast, the discourse of indeterminacy unifies talk about complexity, methodological pluralism, treatment as a process, and the importance of professional autonomy.The idea of professional judgment is interpreted as being a constitutive part of the indeterminacy discourse, which emphasizes the importance of applying a broad range of knowledge and experiences in specific situations.Eraut (1994) defines professional judgment as "the interpretative use of knowledge that implies a practical wisdom, a sense of purpose, appropriateness, and feasibility" (p.49).The strong commitment to ideas of professional judgment expressed by several of the informants in this study was interpreted as being consolidated by an opposition to recent policy initiatives aimed at increasing guidance and regulation of professional work.
In naming the discourses, I was inspired by two studies: the groundbreaking 1970 study by Jamous and Peloille on the changes in the French university-hospital system and the study conducted by Robinson (2003) in which she uses the concepts of technicality and indeterminacy, derived from Jamous and Peloille (1970), as opposing theoretical constructs.Robinson describes the concept of technicality as those aspects of professional work that can be routinized or "programmed" (pp.593-594).In the present study, scientific knowledge and the standardization of professional work are interpreted as important elements in the professionals' talk about practice, thereby leading to the naming of the discourse of technoscience.The concept of indeterminacy is described by Robinson (2003) as "those aspects of practice that are based on specialist knowledge, its interpretations and the use of professional judgment" (pp. 593-594).
An important premise in discourse theory is that discourses will make some practices possible while excluding others.In the following two sections, the discourses will be used in analyzing possibilities for and constraints on evidencebased practice in residential child care.The different definitions of evidence-based practice will be an important part of the discussion.

The discourse of technoscience
The discourse of technoscience is the discourse that most clearly makes possible the ideas of evidence-based practice.The importance given to scientific knowledge and standardization corresponds well with the definitions of evidence-based practice that emphasize the use of best available research evidence in making decisions regarding individual patients.One such definition is that put forward by Sheldon, who described evidence-based practice in social care as "the conscientious, explicit and judicious use of current best evidence in making decisions regarding the welfare of service-users and carers" (as cited in Mullen & Streiner, 2006, p. 113).The use of RCT studies in evidence-based practice has been described as the gold standard in deriving knowledge about professional decisions (Sackett, Rosenberg, Gray, Haynes, & Richardson, 1996).This type of research presupposes the comparison between a randomly selected experimental group and a control group.Several programs in the Norwegian child welfare system have been developed based on the findings of RCT studies (Backe-Hansen, 2009).As Knut pointed out, these programs usually have higher scores on treatment effectiveness.The belief in treatment effectiveness documented through RCT studies was also evident in the interview with Sigrid.Although she expressed ambivalence toward some aspects of the methodological framework, she seemed likely to accept the new methods if they were shown to be effective.
A unitary practice can be contrasted with a traditional view of social work wherein professional practices are characterized by inconsistency related to reflective and interpretative practices and are dominated by professional discretion and judgment (Barfoed & Jacobsson, 2012).The multiplicity of practices and the existence of a vague theoretical framework have been offered as one explanation why social workers in general have been described as occupying a low status in the professional hierarchy (McDonald, 2003;Molander & Terum, 2008).In an article from 2003, McDonald discusses the capacity of evidence-based practice as a strategy for promoting the goals of social work in the context of reforms inspired by New Public Management.This line of reasoning suggests that a unitary practice based on a scientific framework can contribute to reducing uncertainty in professional decisions and, hence, to increasing the status of the social work profession.Larson (1977) states that education and the employment of an accepted knowledge base can be strategic resources for increasing a profession's status.
A belief in scientific knowledge embodies an important idea within the discourse of technoscience, which can accommodate demands for quality improvement from both the professionals themselves and the policymakers, hence making possible evidence-based practice in residential child care.A long history of insufficiencyas pointed out by Olav-seems to be a constitutive part of the discourse of technoscience, which opens up the field for the development of new approaches.
When I asked Olav about the evidence-based nature of the treatment method he had described, he mentioned that the method required major adjustment before implementation.This indicates that within the discourse of technoscience, constraints on evidence-based practice can be found if a narrow definition of evidence-based practice is applied.It is not enough to implement the new methods uncritically; they must also be adjusted.Hence, they must be subject to the professionals' expertise.This is suitable within the wider definition of evidencebased practice, which also incorporates clinical expertise and the use of discretion (Backe-Hansen, 2009;Gilgun, 2005;Satterfield et al., 2009).The use of some discretion within standardized programs, as emphasized in the interview with Olav, shows that standardization does not necessarily reduce professional discretion or professional uncertainty.The opposite is also possible, as standardized programs can force professionals to make decisions constantly about when to use standardized methods or how to adjust them to particular work environments and clients.
The idea of clinical expertise within the discourse of technoscience is also illustrated in the interview with Knut.He used the term "success rate" to describe a statistically proven improvement in treatment outcome measured at evidence-based institutions.Although Knut was aware of this finding, he described his continued skepticism of the method's effectiveness.Only when research evidence was combined with his practical experience did he express enthusiasm for the method.
In summary, the discourse of technoscience not only makes possible the use of current best research evidence and outcome measurement but also includes ideas concerning professional expertise.Based on this discussion, I argue that the discourse of technoscience has the potential to incorporate ideas of evidence-based practice.However, to account fully for the constraints related to professional ambivalence and the need for discretion within the methodological framework, a wide definition of evidence-based practice is needed.

The discourse of indeterminacy
Tensions between standardization and professional judgment in social work practice mark the frontiers of the two identified discourses.In the discourse of indeterminacy, professional judgment is interpreted as an important idea that involves the combination of formal knowledge and practical experience in professional practice.The discourse of indeterminacy allows for complex understandings of professional practice, regarding both the knowledge base and the objectives.The informants' insights into certain situations-for example, Alf's description of how to improve the overall life situations of juveniles and Nina's comment that "substance abuse is often merely a symptom of other challenges"-call for both complex solutions and evaluation procedures.
The importance of autonomy and discretion when making professional judgments about individual juveniles was an oft-mentioned topic during the interviews.Generally, the informants who made use of the discourse of indeterminacy highlighted the variation in work experience and educational background of professionals working within residential child care in Norway, thereby underscoring the importance of professional autonomy and discretion (rather than the ideas of standardization).
The emphases on professional autonomy and discretion within the discourse of indeterminacy can be not only related to the nature of child welfare work (as illustrated in the interviews with Sigrid and Alf), but also interpreted as important elements in being a "professional" and in strengthening professional status.According to Freidson (2001), there are two crucial ideas underlying professionalism: the idea that certain work is so specialized that it needs special training and experience and the idea that certain tasks cannot be standardized or rationalized.The most notable aspect of Freidson's theory is the equal importance placed on experienced-based knowledge and theoretical knowledge (Freidson, 2001).Jamous and Peloille (1970) argue that a profession's status is dependent on the profession's capacity to maintain "indeterminacy" in its practice.According to Jamous and Peloille, there will always exist an inverse relationship between technicality (routinization) and indeterminacy in professional work; hence, if one increases, the other will decrease.
The interview with Therese illuminates the tensions between standardization and professional judgment within professional practice.She expressed her concerns about the new methods and programs (e.g., the policy effort to standardize treatment periods) and called for counter-documentation instead.Within the discourse of indeterminacy, the standardization of treatment will be problematic because it leaves few possibilities for exercising professional judgment and making adjustments in fulfilling the needs of individual juveniles.Hence, if a narrow definition of evidencebased practice is used-implying the implementation of standardized methods and manuals derived from RCT studies-the discourse of indeterminacy will be a constraint on evidence-based residential child care.It can also be argued that the discourse of indeterminacy can make possible evidence-based residential child care if a wide definition of evidence-based practice-that is, one that integrates research evidence and professionals' expertise-is considered.

Conclusion
In the present article, I show opposing professional discourses in residential child care and argue that the possibilities for and constraints on evidence-based practice can be related to how evidence-based practice is defined.I argue that a wide and inclusive definition of the concept of evidence-based practice is as a precondition for encompassing the different elements in the two identified professional discourses.This empirical finding partly supports the theoretical discussion presented by Backe-Hansen (2009) The present article's findings highlight the importance of engaging in constant reflection on the content of evidence-based practice and how it can be related to tensions in professional practice.Further research should elaborate on what to include in a wide definition of evidence-based practice in residential child care and the relationship between the different elements.
This article focused on professional discourses as one source for discussing the possibilities of evidence-based residential child care.An alternative approach would have been to explore different institutional factors and their influence on the possibilities of evidence-based practice.Although this alternative approach was not within the scope of this article, it generates questions for further research.
If you are going to work with people, you have to care, and you have to use your personality.You cannot just use what you have learned to do a good job.
, wherein she argues for a wide definition of evidence-based practice in child welfare work.Whereas the results of the present study primarily point to the importance of the combination of scientific knowledge and professional judgment in a wide definition, Backe-Hansen also emphasizes what the social worker learns from the client and what the client brings to the meeting.